The offsetting advantage is that if the parametric structure we impose with the regression is not too inappropriate, our statistical power along certain dimensions should be enhanced. In particular, if we are interested in doing the analysis over very short intervals of time–e.g., to check the stability of our estimates—the regression approach may be especially useful.
Table 11 summarizes the results. In Panel A, we present the coefficients on the coverage and size variables from cross-sectional regressions run each year over the 14 years 1979-1992.40 We also aggregate the annual information in two different ways. First, we calculate Fama-MacBeth (1973) time-series averages of the coefficients. Second, we run a giant pooled regression with year dummies. Not surprisingly, this latter approach tends to produce point estimates almost identical to the Fama-MacBeth method, but higher t-statistics.
|Pooled w / Year|
All the evidence in Panel A points to a consistent negative effect of analyst coverage on a stock’s serial correlation. Of the yearly coefficients, 13 out 14 are negative, the majority significantly so. The Fama-MacBeth and pooled estimates are strongly significant. The point estimates for size are also negative, but statistically insignificant.
In Panel B, we modify the specification by adding an interaction term, given by log(l+ANALYSTS)*log(SIZE). This is motivated by our evidence in Table 5 that the importance of analyst coverage is decreasing in firm size. The cross-sectional regressions bear out this finding. The coverage and size terms increase in magnitude relative to Panel A (the size term is now statistically significant) and the interaction term is positive, as expected, implying that the negative influence of coverage on serial correlation becomes weaker for larger firms.